Fuzzy Matching for Sentiment Categorization of Contentwise Tweets

Abstract

The evolution of social media platforms has generated immense data that is dynamic, heterogeneous and most often wide spread within the World Wide Web (WWW). This has necessitated researchers to explore techniques for sentiment analysis and classification. Fuzzy pattern matching technique has been widely used to solve decision-making problems that relate to similarity matching. This study utilizes fuzzy matching capabilities for sentiment categorization of contentwise tweets at sentence level using a combination of its matching techniques. Word sentiment polarity and their co-occurrence for pattern matched tweets are further explored. As future works the study will investigate explore automated ‘fuzzy’ sentiment classification for streaming tweets within a given context.

Keywords: fuzzy pattern matching, contentwise tweets, sentiment classification, sentiment polarity